Development of an intelligent electrolytic in process dressing (ELID) grinding system 2

His experimental study suggests that monitoring of the dressing force helps to decide on the time needed for dressing, which will reduce the chances of excessive dressing of the grinding

Chapter 2

Literature review

In-process dressing of grinding wheel has been under research interest for several years

A lot of studies were carried out by many researchers on this subject area ELID is one of the methods of in-process dressing In order to understand its behavior and the field of applications a lot of research works were done previously In this chapter comprehensive discussion will be given on different in-process dressing as well as on ELID grinding In the first section of this chapter various methods of in-process dressing are discussed extensively Previous reports and publications related to ELID grinding and its application are included in the second section of this chapter Furthermore, publications related to grinding wheel monitoring, wheel wear detection and dressing power control are also presented Finally some concluding remarks on the literature study are discussed briefly

2.1 Different dressing/truing techniques

Several methods of in-process dressing of grinding wheels are available as mentioned in the previous chapter Researchers are trying to explore those technologies to investigate their feasibility In the following subsections literatures on all the major in-process dressing techniques are discussed in details

2.1.1 Study on mechanical contact dressing

Albert J Shih [3] studied rotary diamond truing and dressing technique for vitreous bond

wheels for ceramic grinding The author investigated the effects of truing parameters like speed ratio (truing disk surface speed over wheel disk surface speed), feed, overlap ratio (width of contact over truing disk lead), etc on tangential truing force, roundness error of the wheel, surface roughness of the work-piece It was found that a wide range of surface finish, roundness can be generated by varying the truing parameters, which is advantageous because grinding result can be tailored by adjusting truing parameters

Nakagawa et al [4] used stationary dressing stick for in-process dressing of the wheel in

creep feed grinding It was found that normal grinding force is much smaller when process dressing was used The stock removal rate was also found to be higher when used this method of in-process dressing

in-I Inasakai [5] examined the dressing force for resinoid bonded grinding wheels by means

of dynamometer His experimental study suggests that monitoring of the dressing force helps to decide on the time needed for dressing, which will reduce the chances of excessive dressing of the grinding wheel

A.A Torrance et al [6] developed a model for dressing of grinding wheel by diamond

blade or a single point diamond tool The model successfully predicts the topography of the grinding wheel Grinding forces in surface grinding and the surface roughness of the

ground workpiece can also be predicted with good accuracy by this model in combination with abrasive wear model

2.1.2 Electro-thermal method of dressing/truing

K Suzuki et al [7] and T Nakagawa et al [8] developed and studied EDM and WEDM

methods for in-process dressing of the metal bonded grinding wheel It was found that WEDM method produces less electrode wear thus to provide more precise truing and dressing of the grinding wheel whereas die sinking EDM causes more electrode wear Due to the high wear of the block electrode in die sinking case, the electrode has to be machined periodically to maintain the form accuracy Later this group developed a grinding center equipped with their on-machine wire electro-discharge truing/dressing unit for metal bonded diamond grinding wheel They have found it to be very much efficient for different ceramic parts grinding

Parametric study on EDM dressing was carried out by Xiankui Wang et al [9] They

found that truing efficiency depends not only on the electrical parameters of the EDM power but also on wheel speed, wheel eccentricity etc The truing efficiency reduces drastically with the wheel speed after certain threshold level The eccentricity of the wheel plays a positive role on the truing efficiency; the efficiency increases as the eccentricity of the wheel increases

Yan Wang et al [10] proposed dry EDM method of dressing and truing of the grinding

wheel, which helps to reduce the chances of electrolytic corrosion caused by electrolytic

current flowing through water encountered in conventional EDM dressing/truing They compared experimentally the dressing efficiency between the proposed method and the conventional single point mechanical dressing It was found that dry EDM method of wheel dressing is advantageous over the conventional dressing/truing method for metal bonded grinding wheels

Another experimental study for optimization of EDM dressing/truing of metal bonded

CBN grinding wheel was carried out by N Ortega et al [11] It was found that the larger

electrode size gives better stability for truing and dressing process The study also shows

an asymptotic trend in grit protrusion as the operating time of the dressing process increases It was also found in the study that the electro discharged wheel produces much reduced grinding force as compared to the fresh new grinding wheel

Brian K Rhoney et al [12] studied WEDM (wire electro discharge machining) truing/

dressing of metal bonded diamond wheel for silicon nitride machining The study shows WEDM trued/dressed grinding wheel produces significantly (20% to 40%) less grinding force than the single point diamond trued wheel This study also concludes that tool life for the WEDM trued wheel is longer

J A Sanchez [13] developed a new method of EDM based wheel truing method where

the electrode is single point and the movement of the electrode is programmed in such a way that the thickness of the material removed in each dressing pass are constant This

new method removes the manufacturing cost of complex shaped electrode as proposed by the other researchers and it can produce virtually any shape of wheel profile

2.1.3 Laser technology for dressing/truing

X-Z Xie et al [14] used Nd: YAG laser to dress the resin bonded super abrasive wheel

The research focused on studying the variation of ablation crater depth by varying pulse power density It was observed that pulse power density directly influences ablation crater depth They also suggested method of selecting feasible laser parameter for which

a favorable surface topography can be obtained It was concluded in the paper that acousto-optic Q-switched Nd:YAG pulse laser is more suitable for dressing resin-bonded superabrasive grinding wheels than corundum block and CW laser dressing on surface topography

M J Jackson et al [15] studied the effect of high power laser dressing for vitrified

grinding wheels High power laser was also used to remove the clogged chips from the surface of the grinding wheel The change in the properties of the bonding between grinding grains after laser dressing allows the grains to adhere to the surface just long enough so that, after the grits become dull, they may get detached from the surface of the grinding wheel and expose new sharp particles below them thus introducing self sharpening effect in the grinding wheel It was observed in the study that laser dressed grinding wheel performance is equivalent to that of conventionally cleaned and dressed grinding wheel

2.1.4 Electrochemical method of dressing

One of the ways of doing electrochemical dressing of the grinding wheel is the ELID grinding This method of dressing has several important advantages over other dressing methods as described in the table 1.1 of the chapter 1 and this technology is the leading the research area of in-process dressing of super abrasive grinding wheels Therefore its application and other research works on this area will be discussed in the next section separately

In this section another method of Electrochemical dressing will be discussed, which is

known as Electrochemical in-process controlled dressing D.Kramer et al [16] introduced

this method of in-process dressing The principle of operation of this technique is quite similar to that of ELID grinding The main difference is that, this method does not produce any insulating layer during the electrolytic dressing of the wheel In this process the metal bond is just dissolved and thus the new sharp grains protrude out Another interesting feature of this technology is that it uses force feed back to control the dressing current The authors also discussed the performance evaluation of the process It was

found that the average protrusion of the grit was around 100-120% of the grit diameter

The control of the current flow causes the wheel to be worn out less and also the technique eliminates the non-productive dressing time

2.1.5 Miscellaneous methods

Some studies propose other methods of grinding wheel dressing/truing which cannot be

categorized directly under any of the above mentioned methods C.Zhang et al [17]

introduced a combined method of vitrified wheel dressing where laser is used as an assisting medium and actual dressing is carried out by the diamond dresser The laser power softens or even sometime melts the vitrified bonding material thus to facilitate the removal of the bond material by the diamond dresser It was found that truing efficiency

is much higher and wear of the dresser is much less in this case compare to the conventional mechanical contact dressing method The truing/dressing force is also much

reduced in the case of laser assisted dressing In their next study [18] experimental

investigation on the wear rate of the dresser was studied, which was found to be much smaller compared to the mechanical dressing The wear nature in this case is progressive (no trace of fracture wear was found on the worn out dresser) in nature They also evaluated the depending factors of the wear ratio (namely laser power, depth of cut and feed rate) of the dresser for laser assisted diamond dressing

Mingxia et al [19] developed a vibration assisted abrasive belt truing/dressing method for

super abrasive wheel The vibration medium helps to improve the truing and dressing efficiency Super abrasive wheel dressed by this method produces better ground surface than if dressed by other technology (such as Electro discharge method, mechanical contact method.)

ECDM (Electro Chemical Discharge Machining) proposed by M Schopf et al [20] is

another method of dressing super abrasive wheels It is a hybrid process combining both

electrochemical dressing [16] and EDM truing for the grinding wheels The truing

efficiency of the grinding wheel is higher in this case when compared to conventional

truing method The higher protrusion of the sharp grits result in reduced grinding force Furthermore this truing and dressing method produces better surface roughness on the ground cermets workpieces than conventional mechanically dressed/trued grinding wheels

W.K Chen et al [21] proposed the loose abrasive dressing/truing method for resin bond

ultrafine diamond cup wheels for grinding spherical end faces of fibre connectors The truing using a cast iron profiled plate with diamond pastes as lapping media produces a good wheel profile with little wheel damage the loose abrasive dressing was found to be able to produce a much better grit protrusion This method is more suitable for the wheels with finer grit sizes This method of wheel dressing also improves the grinding performance significantly

Ohmori et al [22] also proposed a hybrid method of dressing and truing of ultra fine

grinding wheels The process combines electrolytic in-process dressing (ELID) and electro-discharge truing (ED truing) in one setup It was found that after using this hybrid technology the surface quality, precision of the workpiece and the process stability improved a lot when compared to the conventional grinding without ED truing and ELID

2.2 ELID Grinding

The best known method for super abrasive grinding wheel dressing is electrolytic

in-process dressing (ELID) grinding which was first proposed by Murata et al [23] and later established by Ohmori et al [24] back in 1990 The important feature of this ELID

grinding is that it can be adapted to any existing CNC machine with a very little significant modification The basic principle of ELID grinding has already been discussed

in the first chapter of this report In this section study on different types of ELID grinding, research on fundamental investigations on ELID, and different experimental and development works on ELID grinding have been discussed in details

2.2.1 Preliminary studies on ELID grinding & its classifications

Murata et al [23] are the pioneers of ELID grinding as mentioned earlier They proposed

a current controllable power supply for electrolytic in-process dressing of the metal bonded grinding wheel in order to achieve efficient grinding condition, back in 1985 In their study, the grinding condition was constantly monitored to adjust dressing power by introducing a simple ON-OFF controller Murata et al observed that the grinding stability improved significantly when controlled electrolytic in-process dressing is applied In

1990, Ohmori et al [24] popularized ELID grinding by applying it for grinding of silicon

wafers, however this time monitoring of the grinding condition to adjust the ON-OFF

cycle of the ELID power supply was ignored unlike Murata et al [23] work In their

study Ohmori et al investigated the pre dressing current profile and explained the phenomenon of decreasing trend of the current with dressing time During grinding the grinding force was found to be very stable and continuous because of the continuous protrusion of the sharp grits They also investigated the effect of feed speed on the surface roughness of the ground sample The effect is less significant in the case of ELID grinding, and mirror surface can be achieved at higher feed speed as well

Researchers proposed three major classifications of ELID grinding, based on the materials to be ground and application method However the basic principle of dressing

of the wheel remains the same The various methods of ELID grinding are as follows:

1 Electrolytic In-Process Dressing (ELID 1)

2 Electrolytic Interval Dressing (ELID 2)

3 Electrode-less ELID (ELID 3)

Electrolytic In-Process Dressing (ELID 1)

This method of ELID grinding is mostly used In this case the wheel is dressed

continuously in-process [24] Figure 2.1 shows the basic mechanism of ELID 1 grinding

method A cathode electrode covers usually 1/4th to 1/6th of the cutting edges of the wheel whereas the other exposed portion of the wheel takes part in the cutting action The gap between the electrode and the wheel is maintained to be 0.1mm to 0.3mm

Fig 2.1: ELID 1 grinding principle [24]

Electrolytic Interval Dressing (ELID 2)

In order to carry out precision grinding on the small holes on ceramic materials Zhang et

al [25] introduces ELID 2 method In this method the dressing of the grinding wheel is

carried out intermittently in contrast to ELID 1 method The interval time between two successive dressings depends on pre-decided grinding force limit and other grinding conditions Figure 2.2 shows the schematic illustration of the interval ELID dressing and grinding

Fig2.2: ELID 2 grinding principle [25]

Qian et al [26] also investigated ELID 2 method to grind the internal surface of

cylindrical objects They found that pipe electrode performs better than other shapes of electrodes for dressing the wheel in the case of ELID 2 They also achieved mirror internal surface on an ordinary grinding machine by applying ELID 2 grinding

Electrode-less ELID (ELID 3)

ELID 2 system has an inherent drawback of extra dressing time, for which the grinding

efficiency has to be sacrificed In order to overcome this problem Qian et al [27]

introduced ELID 3 system In ELID 3 system the cathode electrode is removed by the workpiece itself as shown in figure 2.3 However this technique cannot be used for non-conductive material like ceramics, glass, silicon, etc In this grinding method authors used meta-resin bonded wheel The metal is dissolved because of the electrolysis, which

Fig2.3: ELID 3 method [27]

causes the grits to be held by soft resin only This makes the grit and workpiece contact

more elastic and better surface can be achieved as described by the authors [27]

2.2.2 Fundamental study on ELID grinding

Ohmori et al [28] investigated the non linear behavior of electrolytic in-process dressing

It was found in their study that the dressing current reduces over the time during dressing because of the formation of insulating oxide layer This formation of the layer varies with the bond material as described in their study It was also observed that the wheel producing thinner layer experiences lower cutting force because of the higher dressing rate The effect of the types of the power source on the dressing current variation was also studied Dressing current reduces at the highest rate if pure DC power supply is used AC power source has the minimum rate of reduction in the dressing current and for the case of pulsed DC power supply the dressing current reduction rate is in between the above two

pre-Bifano et al [29] developed a model of effective film growth rate by assuming the film

removal rate to be constant with constant load According to the theory the rate of growth

of the film is governed by the following equation:

(2.1)

dt = t

where,

x is the film thickness at time t

a is the electrolytic constant

Lim et al [30] studied the fundamental mechanism of ELID grinding, where they used BK7 glass as the work-piece material They observed non-stability in the grinding force because of the formation and breakage of the insulating ELID layer which is evident throughout grinding The behavior of the insulating layer was studied by comparing the phase relations between the grinding forces and current The observation showed as the dressing current increased, the tangential force increased and the normal force decreased

It was concluded that the oxide layer is not fully worn out but breaks suddenly after reaching a certain condition The condition of breakage of the insulating layer was not explained in the study; however the factors influencing layer breakage were identified and they are the thickness of the insulating layer, the contact area between the workpiece and the wheel, and the feed rate It was also found in their study that a higher current duty ratio produces lesser grinding force and better surface finish because of higher wheel dressing, however the tool wear ratio is more for high duty ratio dressing They also found a threshold feed speed for ELID grinding which is 400mm/min, though it may vary for different work-piece and bond materials

Study on the grinding wheel wear was done by Fathima et al [31] They proposed a

phenomenon that the active grits of grinding during ELID are held by the metal oxide matrix which is softer than the actual metal bond As a result the grit depth of cut in ELID

is reduced and the grinding quality is improved However no quantitative model was proposed for the measurement for reduction in grit depth of cut It was also observed in their study that shorter Ton time is preferred for rough grinding with coarser wheel

whereas longer Ton time gives better surface quality for fine grinding with finer grit wheel

Fathima et al [32] also proposed a grinding force model for ELID grinding Their model established the fact that ELID grinding produces much lower grinding force because of the insulating oxide layer produced during in-process dressing The simulated grinding force and the experimental grinding force were found to be in very good agreement in their study

An important modeling on the electrolysis of ELID was proposed by Hong Chen et al

[33] where authors considered the existence of non conductive diamond particle on a

conductive metal matrix It was found in their study that the current concentration becomes maximum at metal diamond interface; hence the metal dissolution rate Their study also showed that metal dissolution rate increases with higher diamond concentration and lower particle size Finally they concluded that to achieve better performance in ELID the rate of exposing new diamond particle must match the rate of diamond wear; however no model has been proposed for quantifying this phenomenon In order to maintain same metal dissolution rate applied electric field should be lower for higher diamond concentration tool and higher for lower diamond concentration tool

Hong Chen et al [34] further enhanced their model by considering the three dimensional orientation of the diamond particle on a metal matrix The same phenomenon of increasing electric field concentration at the metal diamond boundary was observed for

the 3-D case Furthermore they investigated the protrusion effect of the diamond particle

on the electric field concentration It was found that the field concentration effect at the metal diamond boundary decreases rapidly as the protrusion of the grits increases and the electric field becomes more uniform However authors did not consider the effect of ELID layer on the protrusion of the diamond grits which is the actual case

2.2.3 Study on control of ELID grinding

There are very few studies on the control of ELID grinding In this section all those studies will be discussed comprehensively

Rich Boland et al [35] developed a current feed back system to control the feed override during grinding It was observed in their study that dressing current during ELID grinding follows a non uniform pattern At the initial stage of the grinding the current is very high because of the absence of oxide layer and at the final stage of grinding the current becomes very low as the oxide layer grows sufficiently The authors studied the frequency spectrum of the current to identify which frequency change is dominant during ELID grinding It was observed in their study that at 49Khz there is a definite change in the frequency spectrum of the dressing current because of the above mentioned phenomenon Based on previously machined parts a threshold was set for the amplitude

of the frequency spectrum at 49 KHz A control loop was then developed to override the feed speed during grinding; which means the feed speed slowed down if the spectrum was above threshold value and it increased if the spectrum was below threshold value By applying this control technique authors achieved much uniform current profile for the

whole grinding cycle, however the effect of uniform dressing current on the final quality

of the ground workpieces was not investigated in the paper

Eun-Sang Lee [36, 37] studied ELID grinding of die steel using optimum dressing

control It was shown in their study that insulating layer thickness decreases as the gap between the electrodes increases In conventional ELID grinding the layer formation is non uniform which is not favorable for constant condition grinding However it was experimentally observed from authors’ study that the layer thickness can be controlled by varying the dressing current The relationship between the layer thickness and dressing current is expressed as follows,

where,

to is the layer thickness

and I is the dressing current

After applying the dressing current control the layer thickness was found to be uniform through out the grinding process The basic algorithm proposed by the authors for dressing current control is as follows:

Firstly the regular current limit, initial peak dressing current, ton and toff are set As the grinding starts the real current is continuously monitored and if there is any deviation from the regular current setting then the peak dressing current is adjusted accordingly The experimental results show that this control algorithm reduces the grinding force and improves the surface roughness significantly as compared to conventional ELID grinding

M Ashizuka et al [38] introduced a new concept of grinding wheel truing by controlled electrolytic in-process dressing Authors assumed in their study that the ELID layer is very soft and brittle in nature and shall wear out easily as it interacts with the workpiece

In this study authors used an inductive sensor to measure the metal bond profile of the wheel which gives a measurement of wheel non uniformity After observing the measurement the high portion of the wheel was dressed more by applying higher pulse duty ratio which causes more insulating layer formation at the high zone of the wheel, however this excess layer shall be worn out as the wheel interacts with the workpiece Although the concept of wheel truing by controllable wheel dressing is very new and interesting, authors ignored the convolution effect between the dresser electrode and the wheel which shall be addressed later in this thesis

2.2.4 Study on various application of ELID grinding

The feasibility of ELID grinding has been investigated for wide range of hard and brittle materials including ceramics, optical glass, hardened steel, silicon etc In most of the cases the results were very much optimistic In this section the studies related to the applications of ELID grinding shall be discussed comprehensively

ELID on Ceramics

Ohmori et al [39] carried out an experimental study to investigate the performance of

ELID grinding on structural ceramics particularly for silicon nitrides Two different types

of silicon nitrides were used in this purpose namely a sintered reaction-bonded silicon nitride (SRBSN), and a cast-and-sintered silicon nitride (Si3N4) The effects of cutting